tierney {anomaly}R Documentation

tierney

Description

Transforms the data X by centring and scaling using X_{ij}^{'} = \frac{X_{ij}-μ_{ij}}{σ_{ij}} where μ_{ij} and σ_{ij} are robust quantile based sequential estimates for the mean and standard deviation of each variate (column) X_{i} of X calculated up to time j. The estimates μ_{ij} and σ_{ij} are calculated from sequential estimates for the median and inter-quartile range developed by Tierney et al (1983). This method is the default value for the transform argument used by the scapa.uv function.

Usage

tierney(X, burnin = 10)

Arguments

X

A numeric matrix containing the data to be transformed.

burnin

Specifies the period used to stabalise the quantile estimates. The default value is 10.

Value

A numeric matrix containing the transformed data.

References

Schruben L, Singh H, Tierney L (1983). “Optimal Tests for Initialization Bias in Simulation Output.” Oper. Res., 31(6), 1167–1178. ISSN 0030-364X, doi: 10.1287/opre.31.6.1167, https://doi.org/10.1287/opre.31.6.1167.

Examples

library(anomaly)
data(machinetemp)
attach(machinetemp)
plot(temperature)
temperature<-tierney(temperature,burnin=4305)
plot(temperature)

[Package anomaly version 4.0.1 Index]